Presentation

<
31. Neuro-Symbolic Transformation of Architectural Facades into Their Procedural Representations
SessionPosters: Geometry & Modeling
DescriptionThis work introduces a neuro-symbolic transformer model that automates the conversion of segmented facade images into procedural definitions using a custom split grammar. It enhances architectural design flexibility, allowing easy adjustments and variations, and sets a new standard in facade design by merging procedural generation with neuro-symbolic learning.

INVITED TO THE FIRST ROUND OF THE STUDENT RESEARCH COMPETITION
Contributors
Aleksander Płocharski
Warsaw University of Technology
IDEAS NCBR
Jan Swidzinski
IDEAS NCBR
Joanna Porter-Sobieraj
Warsaw University of Technology
Przemyslaw Musialski
New Jersey Institute of Technology (NJIT)
IDEAS NCBR
Event Type
Poster
TimeSunday, 28 July 20249:00am - 5:30pm MDT
LocationMile High Pre-Function Area
Session Time & Location
Sunday, 28 July 20249:00am - 5:30pm MDTMile High Pre-Function Area
Monday, 29 July 20249:00am - 5:30pm MDTMile High Pre-Function Area
Tuesday, 30 July 20249:00am - 5:30pm MDTMile High Pre-Function Area
Wednesday, 31 July 20249:00am - 5:30pm MDTMile High Pre-Function Area
Thursday, 1 August 20249:00am - 5:30pm MDTMile High Pre-Function Area
Registration Categories
Full Conference
Full Conference Supporter
Experience
Exhibitor Full Conference
Exhibitor Experience
Next PresentationNext Presentation32. O, What an Iridescent Web We Weave: Rendering Physically Inspired Spider Webs for Visual Effects